Background Biological disease-modifying anti-rheumatic drugs (bioDMARDs) are generally used in combination with synthetic DMARDs. However, data derived from several registries suggest that bioDMARDs are used in monotherapy in up to one-third of RA patients.

Objectives To compare the treatment maintenance of bioDMARDs started in monotherapy or in combination with synthetic DMARDs in RA patients from the SCQM registry.

Methods Demographic and disease characteristics of all RA patients treated with bioDMARDs were retrieved from the SCQM Registry. Monotherapy and combination therapy were defined according to whether bioDMARDs were introduced with or without concomitant synthetic DMARDs, respectively. Treatment maintenance was measured by the time from start to discontinuation of the bioDMARD. Treatment courses of patients who started abatacept, adalimumab, certolizumab, etanercept, golimumab, infliximab, or tocilizumab, who were not lost to follow-up immediately and provided complete covariate information were included. A Cox proportional hazard model with bioDMARD as stratification variable was applied and adjusted for the following covariates: type of therapy (mono versus combination), age, sex, BMI, smoking, number of previous bioDMARDs, number of bioDMARD options, disease duration, seropositivity, disease activity, and functional disability.

Results 3853 treatment courses with bioDMARDs in 2937 patients were retrieved. BioDMARDs were prescribed in monotherapy in 1020 treatment courses (27%). Certolizumab, etanercept and tocilizumab were the bioDMARDs started most frequently in monotherapy (17/42 (40%), 381/1092 (35%), and 78/225 (35%) treatment courses, respectively), whereas infliximab was the least used in monotherapy 85/646 (13%). The addition of synthetic DMARDs during the course of monotherapy was ascertained in 11% of the monotherapy cases, and a switch to monotherapy (i.e. discontinuation of all synthetic DMARDs) prior to the end of treatment was ascertained in 11% of combination therapy cases. A total of 2231 of the 3853 treatment courses met the requirements for inclusion into the analysis of bioDMARD treatment maintenance. A total of 1421 of these constituted observed treatment discontinuation events. Age (P=0.03), BMI (P=0.002), disease duration (P<0.0001), number of previous bioDMARDs (P=0.006), number of bioDMARD options (P=0.003), and type of bioDMARDs (P=0.0005) were significantly different in patients treated in monotherapy as compared to those on combination therapy, whereas sex distribution, DAS28, HAQ, and sepositivity were not different in the two groups. The estimated hazard ratio for mono versus combination therapy was 1.14 with a 95%>CI of [1.004, 1.292]. Furthermore, bioDMARD maintenance was shorter with increasing numbers of previously received bioDMARDs, higher disease activity (DAS28ESR), and higher functional disability (HAQ) and longer with increasing disease duration. Higher numbers of treatment options were associated with higher hazard rates for bioDMARD discontinuation, i.e. shorter maintenance.

Conclusions The results of this observational study of RA patients suggest that bioDMARD retention is slightly reduced when used in monotherapy.

Acknowledgements This study was supported by an unrestricted grant from Roche

Disclosure of Interest C. Gabay Grant/research support from: Roche, Abbott, Pfizer, MSD, Consultant for: Roche, Abbott, MSD, Pfizer, BMS, M. Riek: None Declared, A. Scherer: None Declared, A. Finckh Grant/research support from: Roche, Abbott, Pfizer, BMS, Consultant for: Roche, Pfizer